CN112580108B - Signature and seal integrity verification method and computer equipment - Google Patents
Signature and seal integrity verification method and computer equipment Download PDFInfo
- Publication number
- CN112580108B CN112580108B CN202011434668.9A CN202011434668A CN112580108B CN 112580108 B CN112580108 B CN 112580108B CN 202011434668 A CN202011434668 A CN 202011434668A CN 112580108 B CN112580108 B CN 112580108B
- Authority
- CN
- China
- Prior art keywords
- seal
- signature
- verified
- target
- chapter
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000012795 verification Methods 0.000 title claims abstract description 50
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000001514 detection method Methods 0.000 claims abstract description 87
- 230000000875 corresponding effect Effects 0.000 claims description 62
- 238000012549 training Methods 0.000 claims description 32
- 238000013145 classification model Methods 0.000 claims description 14
- 238000010801 machine learning Methods 0.000 claims description 9
- 230000002596 correlated effect Effects 0.000 claims description 2
- 238000012545 processing Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 4
- 238000013527 convolutional neural network Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000012015 optical character recognition Methods 0.000 description 3
- 230000008520 organization Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000012550 audit Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 238000007789 sealing Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/60—Protecting data
- G06F21/64—Protecting data integrity, e.g. using checksums, certificates or signatures
Abstract
The embodiment of the application discloses a signature and seal integrity verification method, computer equipment and a computer storage medium, which are used for rapidly and automatically verifying a signature and a seal so as to improve the verification efficiency of the signature and the seal. The embodiment of the application comprises the following steps: the method comprises the steps that computer equipment detects a document to be verified according to a target detection model which is trained in advance, determines signature areas and seal areas in a section to be verified of the document to be verified, judges whether the number of signature texts in all signature areas in the section to be verified is smaller than the number of signature identifiers in the target section, and if yes, determines that the signature of the section to be verified is incomplete; judging whether the number of the stamps in all stamp areas in the chapter to be verified is less than the number of the stamp marks in the target chapter, if so, determining that the stamps of the chapter to be verified are incomplete. Therefore, the automatic verification of the integrity of the signature seal of the document is realized, the verification of the signature seal is not required to be manually executed, and the verification efficiency of the signature seal is improved.
Description
Technical Field
The embodiment of the application relates to the field of data processing, in particular to a signature and seal integrity verification method and computer equipment.
Background
At present, paperless office and office automation systems (office automation, OA) are popular, but the application of paper documents with entity seals in the financial field is still very wide, especially in the fields of trans-institution information reporting or trust transfer. In the aspect of securities issuance, the sponsoring organization needs to submit mass files such as a stranding instruction, an audit report, a sponsoring report and the like, the number of pages of the files is large, some files even reach thousands of pages, the places needing signature and seal are large, the workload of manually verifying the signature and seal is large, and the working efficiency is greatly limited.
Therefore, a scheme for automatically verifying the signature and the seal is needed to improve the verification efficiency of the signature and the seal.
Disclosure of Invention
The embodiment of the application provides a signature and seal integrity verification method, computer equipment and a computer storage medium, which are used for rapidly and automatically verifying a signature and a seal so as to improve the verification efficiency of the signature and the seal.
The first aspect of the embodiment of the application provides a signature and seal integrity verification method, which comprises the following steps:
acquiring a current document to be verified, and acquiring a pre-constructed target dictionary, wherein the target dictionary comprises the corresponding relation between chapters and sections of the document to be verified and signature identifiers and seal identifiers, and the target dictionary corresponds to the document category of the document to be verified;
Detecting the document to be verified according to a target detection model which is trained in advance, and determining a signature area and a seal area in a chapter to be verified of the document to be verified;
Determining a target chapter corresponding to the chapter to be verified in the target dictionary, judging whether the number of signature texts of all signature areas in the chapter to be verified is less than the number of signature identifiers in the target chapter, and judging whether the number of seals of all seal areas in the chapter to be verified is less than the number of seal identifiers in the target chapter;
If the number of the signature texts is equal to the number of the signature identifications and the number of the seals is equal to the number of the seal identifications, determining that the signature and the seal of the chapter to be verified are complete;
If the number of the signature texts is smaller than the number of the signature identifiers, determining that the signature of the chapter to be verified is incomplete;
And if the number of the seals is less than the number of the seal marks, determining that the seals of the chapter to be verified are incomplete.
A second aspect of an embodiment of the present application provides a computer apparatus, including:
The acquisition unit is used for acquiring a current document to be verified and acquiring a pre-constructed target dictionary, wherein the target dictionary comprises the corresponding relation between chapters and signature identifiers and seal identifiers of the document to be verified, and the target dictionary corresponds to the document category of the document to be verified;
the target detection unit is used for detecting the document to be verified according to a target detection model which is trained in advance, and determining a signature area and a seal area in a chapter to be verified of the document to be verified;
a first determining unit, configured to determine a target chapter corresponding to the chapter to be verified in the target dictionary;
The judging unit is used for judging whether the number of signature texts of all signature areas in the chapter to be verified is less than the number of signature identifiers in the target chapter and judging whether the number of seals of all seal areas in the chapter to be verified is less than the number of seal identifiers in the target chapter;
The second determining unit is used for determining the signature and seal integrity of the chapter to be verified when the number of signature texts is equal to the number of signature identifiers and the number of seals is equal to the number of seal identifiers;
The second determining unit is further used for determining that the signature of the chapter to be verified is incomplete when the number of signature texts is smaller than the number of signature identifiers;
and the second determining unit is also used for determining that the seal of the chapter to be verified is incomplete when the number of the seals is less than the number of the seal marks.
A third aspect of an embodiment of the present application provides a computer apparatus, including:
a processor, a memory, a bus, and an input/output device;
The processor is connected with the memory and the input and output equipment;
the bus is respectively connected with the processor, the memory and the input and output equipment;
The input/output equipment is used for acquiring a current document to be verified and acquiring a pre-constructed target dictionary, wherein the target dictionary comprises the corresponding relation between chapters and sections of the document to be verified and signature identifiers and seal identifiers, and the target dictionary corresponds to the document category of the document to be verified;
The processor is used for detecting the document to be verified according to a target detection model which is trained in advance, and determining a signature area and a seal area in a chapter to be verified of the document to be verified; determining a target chapter corresponding to the chapter to be verified in the target dictionary; judging whether the number of signature texts of all signature areas in the chapter to be verified is less than the number of signature identifiers in the target chapter, and judging whether the number of seals of all seal areas in the chapter to be verified is less than the number of seal identifiers in the target chapter; when the number of the signature texts is equal to the number of the signature identifications and the number of the seals is equal to the number of the seal identifications, determining that the signature and the seal of the chapter to be verified are complete; when the number of signature texts is smaller than the number of signature identifications, determining that the signature of the chapter to be verified is incomplete; and when the number of the seals is less than the number of the seal marks, determining that the seals of the chapter to be verified are incomplete.
A fourth aspect of the embodiments of the present application provides a computer storage medium having stored therein instructions which, when executed on a computer, cause the computer to perform the method of the first aspect described above.
From the above technical solutions, the embodiment of the present application has the following advantages:
In the embodiment of the application, the computer equipment detects a document to be verified according to a target detection model which is trained in advance, determines a signature area and a seal area in a section to be verified of the document to be verified, judges whether the number of signature texts in all signature areas in the section to be verified is less than the number of signature identifiers in a target section, and judges whether the number of seals in all seal areas in the section to be verified is less than the number of seal identifiers in the target section; when the number of the stamps in all the stamp areas is less than the number of the stamp marks in the target chapter, determining that the stamps of the chapter to be verified are incomplete. Therefore, the automatic verification of the integrity of the signature seal of the document is realized, the verification of the signature seal is not required to be manually executed, and the verification efficiency of the signature seal is improved.
Drawings
FIG. 1 is a schematic flow chart of a signature and seal integrity verification method in an embodiment of the application;
FIG. 2 is a schematic flow chart of a signature and seal integrity verification method in an embodiment of the application;
FIG. 3 is a schematic diagram showing an identification effect of a signature area and a seal area by a target detection model according to an embodiment of the present application;
FIG. 4 is a schematic diagram of a signature area and a stamp area in a chapter to be verified in an embodiment of the present application;
FIG. 5 is a schematic diagram of a computer device according to an embodiment of the present application;
FIG. 6 is a schematic diagram of another configuration of a computer device according to an embodiment of the present application.
Detailed Description
The embodiment of the application provides a signature and seal integrity verification method, computer equipment and a computer storage medium, which are used for rapidly and automatically verifying a signature and a seal so as to improve the verification efficiency of the signature and the seal.
Referring to fig. 1, an embodiment of a signature and seal integrity verification method in an embodiment of the present application includes:
101. acquiring a current document to be verified, and acquiring a pre-constructed target dictionary;
The method of the present embodiment is applied to a computer device, which may be a terminal, a server, or the like, capable of performing tasks such as data processing and data analysis. When the computer device is a terminal, the computer device may be a terminal device such as a personal computer (personal computer, PC) or a desktop computer; when the computer device is a server, the computer device can be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server for providing basic cloud computing services such as a cloud database, cloud computing, big data, an artificial intelligent platform and the like.
The computer device obtains the document to be verified to verify whether the signature and the seal in the document to be verified are complete. When the document to be verified is obtained, the document to be verified is scanned or photographed, the document to be verified is converted into the form of an electronic document so that the computer equipment can recognize and read the document to be verified, and the conversion of the document to be verified into electronic data is also beneficial to data transmission and data processing.
In this embodiment, the document to be verified includes one or more chapters, each of which is a portion of the document to be verified, and each of which also includes one or more signatures and seals. And a person manually constructs a target dictionary corresponding to the document category of the document to be verified according to the document category of the document to be verified, and inputs the target dictionary into the computer equipment. Thus, the target dictionary corresponds to a document in the document category to which the document to be verified belongs. Specifically, the target dictionary includes the correspondence between the chapters of the document to be verified and the signature identifier and the seal identifier, that is, the target dictionary describes the signature identifier and the seal identifier included in each chapter, and the signature identifier and the seal identifier may be characters such as letters and numbers to refer to the signature and the seal respectively. For example, the signature identifier may be an identifier such as signature 1 or signature 2, and the stamp identifier may be an identifier such as stamp 1 or stamp 2.
For example, the person constructs a target dictionary from a document of the category "combined cash flow amount", the target dictionary describing a signature and a stamp that should be present in the combined cash flow amount. And when verifying the current certain combined cash flow table, acquiring the target dictionary, and verifying the signature seal integrity of the current combined cash flow table by using the target dictionary.
102. Detecting a document to be verified according to a target detection model which is trained in advance, and determining a signature area and a seal area in a chapter to be verified of the document to be verified;
In this embodiment, the target detection model is trained in advance, and the target detection model can be used to identify and detect the targets of the signature area and the seal area in the document. Because the document to be verified comprises one or more chapters, the target detection model is utilized to carry out target identification on the chapter to be verified in the document to be verified, and a signature area and a seal area in the chapter to be verified are determined.
One signature area in the chapter to be verified corresponds to a location where a signature is required, for example, the chapter to be verified has 5 signature areas, which indicates that the chapter to be verified has 5 locations where a signature is required. And similarly, one seal area in the chapter to be verified corresponds to one position needing to be stamped.
The signature area can be classified into the following categories according to actual conditions: classes with only signature cues without signature text, with signature cues and signature text, with only signature text without signature cues, etc.; the seal area can be divided into the following categories according to actual conditions: only the seal prompt without seal content, seal prompt and seal content, only the seal content without seal prompt and other categories. The signature text refers to signature handwriting of the principal and text corresponding to the signature handwriting (namely name of the principal), the seal content refers to seal patterns and text corresponding to the seal patterns, if the seal is an organization seal, the text corresponding to the seal patterns is an organization name, and if the seal is a name seal, the text corresponding to the seal patterns is a name. It will be appreciated that when the stamp is a name stamp, the signature hint may also be considered a stamp hint.
103. Determining a target chapter corresponding to the chapter to be verified in the target dictionary;
Because the target dictionary describes the correspondence between the chapters of the document to be verified and the signature identifier and the seal identifier, the target chapter corresponding to the chapter to be verified can be determined from the target dictionary. The target chapter corresponds to the chapter to be verified, and may be a chapter name corresponding to the chapter, for example, the chapter of "combined cash flow table" of the document to be verified corresponds to the chapter of "combined cash flow table" in the target dictionary; it may also be the page number of a chapter that corresponds, or otherwise corresponds.
It should be noted that, step 103 may be performed prior to step 102, may be performed after step 102, or may be performed simultaneously with steps 102 and 103, and the order of performing steps 102 and 103 is not limited.
104. Judging whether the number of signature texts of all signature areas in the section to be verified is less than the number of signature identifications in the target section, if so, executing step 105; if not, executing step 106;
After the target detection model identifies the signature areas in the section to be verified, judging whether the number of signature texts in all the signature areas in the section to be verified is less than the number of signature identifiers in the target section.
For example, there are 5 signature areas in the section to be verified, and only 3 signature areas of the 5 signature areas are signed, i.e. only 3 signature texts. Meanwhile, if 5 signature identifiers exist in the target chapter corresponding to the chapter to be verified, it can be determined that the number of signature texts of all signature areas in the chapter to be verified is smaller than the number of signature identifiers in the target chapter.
105. Determining that the signature of the section to be verified is incomplete;
Because the target dictionary is constructed by personnel according to the actual conditions of signing and stamping required by the document, the integrity of the signature seal of the document can be verified by taking the target dictionary as a standard. When the number of signature texts of all signature areas in the section to be verified is less than the number of signature identifications in the target section, the section to be verified is not in accordance with the standard determined by the target dictionary, namely, the section to be verified has partial signature areas which are not signed, and the signature is incomplete.
106. Determining that the signature of the section to be verified is complete;
And if the number of signature texts of all the signature areas in the chapter to be verified is equal to the number of signature identifiers in the target chapter, indicating that the chapter to be verified accords with the standard determined by the target dictionary, and the signature of the chapter to be verified is complete.
107. Judging whether the number of the seals of all seal areas in the chapter to be verified is less than the number of the seal marks in the target chapter, if so, executing step 108; if not, go to step 109;
After the target detection model identifies the seal areas in the chapter to be verified, judging whether the number of seals in all seal areas in the chapter to be verified is less than the number of seal marks in the target chapter.
For example, there are 5 seal areas in the chapter to be verified, and only 3 seal areas of these 5 seal areas are sealed, i.e., the number of seals is 3. Meanwhile, if 5 seal marks exist in the target chapter corresponding to the chapter to be verified, it can be determined that the number of seals in all seal areas in the chapter to be verified is smaller than the number of seal marks in the target chapter.
108. Determining that the seal of the chapter to be verified is incomplete;
When the number of the stamps in all the stamp areas in the chapter to be verified is less than the number of the stamp marks in the target chapter, the chapter to be verified is not in accordance with the standard determined by the target dictionary, namely, the chapter to be verified is provided with partial stamp areas which are not stamped, and the stamp is incomplete.
109. Determining the completeness of the seal of the chapter to be verified;
and if the number of the stamps in all the stamp areas in the chapter to be verified is equal to the number of the stamp marks in the target chapter, indicating that the chapter to be verified accords with the standard determined by the target dictionary, and the stamps of the chapter to be verified are complete.
After obtaining the result of whether the signature and the seal of the chapter to be verified are complete, the result can be output to the user of the computer device, so that the user can know the verification result of the chapter to be verified.
And respectively verifying a plurality of chapters to be verified of the document to be verified according to the steps, so that the signature of the document to be verified and the integrity verification result of the seal can be obtained.
In this embodiment, the computer device detects the document to be verified according to the target detection model trained in advance, determines the signature area and the seal area in the section to be verified of the document to be verified, determines whether the number of signature texts in all the signature areas in the section to be verified is less than the number of signature identifiers in the target section, and determines whether the number of seals in all the seal areas in the section to be verified is less than the number of seal identifiers in the target section, because the target dictionary is constructed by personnel according to the actual conditions of signing and sealing required by the document, the target dictionary can be used as a standard for verifying the signature seal integrity of the document, and therefore, determines that the signature of the section to be verified is incomplete when the number of signature texts in all the signature areas is less than the number of signature identifiers in the target section; when the number of the stamps in all the stamp areas is less than the number of the stamp marks in the target chapter, determining that the stamps of the chapter to be verified are incomplete. Therefore, the automatic verification of the integrity of the signature seal of the document is realized, the verification of the signature seal is not required to be manually executed, and the verification efficiency of the signature seal is improved.
An embodiment of the present application will be described in further detail below on the basis of the foregoing embodiment shown in fig. 1. Referring to fig. 2, another embodiment of a signature and seal integrity verification method according to an embodiment of the present application includes:
201. Acquiring a current document to be verified, and acquiring a pre-constructed target dictionary;
In this embodiment, the target dictionary is constructed according to the standard that the signature and seal are required in the document of the document class to which the document to be verified belongs, that is, the target dictionary describes the signature and seal that the document of the document class to which the document to be verified should have. Thus, after the target dictionary is constructed, multiple documents in one document category may be verified according to the target dictionary.
For example, assuming that the document to be verified includes a combined cash flow table and an accounting statement, and the combined cash flow table and the accounting statement each include a signature and a stamp that must be included, a target dictionary is constructed according to the above criteria, and the target dictionary describes the signature and the stamp that must be included in the combined cash flow table and the accounting statement, resulting in a target dictionary shown in table 1, in which signature 1, stamp 1, signature 2, stamp 2, and the like are a signature identifier and a stamp identifier, respectively.
TABLE 1
The combined cash flow table and the accountant transaction statement in the document to be verified can be respectively used as one section to be verified of the document to be verified, and then a target section corresponding to any section to be verified exists in the target dictionary, for example, the section to be verified 'combined cash flow table' corresponds to the section to be verified 'combined cash flow table' in the target dictionary.
202. Detecting a document to be verified according to a target detection model which is trained in advance, and determining a signature area and a seal area in a chapter to be verified of the document to be verified;
In this embodiment, the target detection model is obtained by training multiple sets of training samples based on a machine learning algorithm, where each set of training samples includes a signature seal type image, that is, multiple signature seal type images including an image with only a signature prompt, an image with only a seal prompt, an image with only a signature prompt and a signature, an image with only a seal prompt and a seal, an image with a signature prompt and a signature text and a seal prompt and a seal, an image with a signature prompt and a seal prompt, an image with a seal prompt and a seal and a signature prompt, and the like are used as training samples, and the initial model of the target detection model trains the multiple signature seal type images based on the machine learning algorithm, when training is performed until convergence conditions of the model are satisfied, the target detection model is obtained.
Therefore, the target detection model trained by the model can be used for carrying out target identification on the signature area and the seal area in the page of the document to be verified, the document to be verified is input into the target detection model, the target detection model is used for identifying the signature area and the seal area in the page, and the area position information of the signature area and the area position information of the seal area are output.
The region position information of the signature region and the region position information of the seal region may be coordinate information of the signature region and coordinate information of the seal region, where the coordinate information is a diagonal point coordinate of the target region marked by a rectangular frame.
For example, as shown in fig. 3, the target detection model identifies the signature area and the seal area respectively, and marks the identified signature area and seal area in the form of rectangular frames, wherein the diagonal coordinates of the rectangular frames are the area position information of the signature area or seal area.
It will be appreciated that, because the target detection model is trained based on the above-described multiple signature seal class images, the target detection model may identify types of signature areas such as signature prompt only, signature and name seal only, and may also identify types of seal areas such as seal prompt only, seal only, and the like.
The machine learning algorithm for training the target detection model can be a machine vision algorithm such as R-CNN, fast R-CNN, YOLO, SSD and the like.
The present embodiment also extracts content in the signature area and the stamp area based on optical character recognition (optical character recognition, OCR) technology. Specifically, the computer device obtains a text detection model that is trained by a text detection algorithm on multiple sets of image training samples. After the target detection model determines the area position information of the signature area and the seal area, inputting an image corresponding to the area position information of the signature area into the text detection model, performing text detection on the image to obtain a signature text and/or a signature prompt text of the signature area output by the text detection model, inputting the image corresponding to the area position information of the seal area into the text detection model, performing text detection on the image, and obtaining seal content and/or seal prompt text of the seal area output by the text detection model.
The text detection algorithm may be CTPN text detection algorithm (DETECTING TEXT IN NATURAL IMAGE WITH connectionist text proposal network) based on convolutional neural network and cyclic neural network, and may also be EAST algorithm (EFFICIENT AND accuracy scene text) or SegLink algorithm.
Each group of training samples of the target detection model further comprises label information for representing the classification result of the signature seal class image, the target detection model can train a plurality of groups of training samples, and the corresponding relation between the label information and the signature seal class image is fitted continuously until the training of the model meets the convergence condition. The trained target detection model can classify the signature seal class image to be identified and output the classification result.
Therefore, the target detection model can classify the signature region and the seal region and output the classification result of the signature region and the classification result of the seal region at the same time of determining the signature region and the seal region.
The label information for representing the classification result of the signature seal type image may be any character, for example, a number "1" as label information of an image having only a signature hint and a signature, a number "2" as label information of an image having a signature hint and a signature and a name seal, and the like.
203. Determining a target chapter corresponding to the chapter to be verified in the target dictionary;
The operation performed in this step is similar to the operation performed in step 103 in the embodiment shown in fig. 1, and will not be described here again.
204. Filling the signature text of the signature area into a filling area indicated by a target signature mark in a target chapter, and filling the seal content of the seal area into the filling area indicated by the target seal mark in the target chapter;
In this embodiment, it may be determined whether the number of signature texts is less than the number of signature identifiers in the target chapter and whether the number of seals is less than the number of seal identifiers in the target chapter by filling the signature texts and seal contents into the target dictionary. Specifically, the signature identifier in the target chapter includes the region position information corresponding to the signature identifier, and the seal identifier in the target chapter includes the region position information corresponding to the seal identifier. And determining a target signature mark and a target seal mark in a target chapter of the target dictionary, wherein the region position information corresponding to the target signature mark corresponds to the region position information of the signature region, and the region position information corresponding to the target seal mark corresponds to the region position information of the seal region. After detecting the signature text of the signature area and the seal content of the seal area in step 202, the signature text of the signature area is filled into a filling area indicated by a target signature identifier corresponding to the signature area in the target chapter, and the seal content of the seal area is filled into a filling area indicated by a target seal identifier corresponding to the seal area in the target chapter.
For example, following the example shown in table 1, in the target chapter "combined cash flow table" of the target dictionary shown in table 1, each signature mark and each seal mark include corresponding area position information, that is, the area position information of the signature mark indicates the position of the signature mark in the page, and the area position information of the seal mark indicates the position of the seal mark in the page. Because the signature area and the seal area in the section to be verified also respectively comprise the area position information, the target signature mark corresponding to the signature area and the target seal mark corresponding to the seal area can be determined. And then filling the signature text of the signature area into a filling area indicated by the target signature mark, and filling the seal content of the seal area into the filling area indicated by the target seal mark. Assuming that the signature area and the seal area in the chapter to be verified are as shown in fig. 4, the target chapter of the filled signature text and seal content shown in table 2 can be obtained.
TABLE 2
205. Judging whether filling areas indicated by all signature identifiers in the target section have filling areas without filling signature texts, if so, executing step 206; if not, go to step 207;
The target detection model detects all signature areas and all seal areas in a page of a section to be verified, when all signature areas have signed names, the target detection model can detect signature texts of all signature areas, signature texts of each signature area are respectively filled into filling areas of corresponding signature identifiers, and then filling areas indicated by all signature identifiers in the target section are filled with the signature texts; if some signature areas are not signed with names, the target detection model can only detect signature texts of the signature areas of the signed names, and fills the signature texts into target chapters, and filling areas, indicated by all signature identifiers in the target chapters, still have filling areas without filling the signature texts.
206. Determining that the number of signature texts is less than the number of signature identifiers, and the signature of the chapter to be verified is incomplete;
If a filling area of unfilled signature texts exists, namely the number of signature texts in the section to be verified is smaller than the number of signature identifiers in the target section, which indicates that part of signature areas in the section to be verified have unsigned names, the incomplete signature of the section to be verified can be determined.
207. Determining that the signature of the section to be verified is complete;
If the filling area of the unfilled signature text does not exist, namely the filling area indicated by all the signature identifiers in the target chapter is filled with the signature text, which indicates that all the signature areas which should sign names in the chapter to be verified are signed with names, the signature of the chapter to be verified can be determined to be complete.
208. Judging whether filling areas indicated by all seal marks in the target chapter are filled areas without filling seal contents or not, if so, executing step 209; if not, go to step 210;
Similarly, when all seal areas are sealed, the target detection model can detect seal contents of all seal areas, seal contents of each seal area are respectively filled into filling areas of corresponding seal marks, and then the filling areas indicated by all seal marks in the target chapter are filled with seal contents; if a part of seal areas are not stamped, the target detection model only detects seal contents of stamped seal areas, the seal contents are filled into target chapters, and filling areas indicated by all seal marks in the target chapters still have filling areas without filling seal contents.
209. Determining that the number of the seals is less than the number of the seal marks, and the seals of the chapter to be verified are incomplete;
if the filling area of the unfilled seal content exists, namely the number of the seals in the section to be verified is smaller than the number of the seal marks in the target section, which indicates that part of the seal area in the section to be verified is not stamped, the seal of the section to be verified can be determined to be incomplete.
210. Determining the completeness of the seal of the chapter to be verified;
If the filling areas of the unfilled seal content do not exist, namely the filling areas indicated by all seal identifications in the target chapter are filled with the seal content, and the seal content indicates that all seal areas which should be sealed in the chapter to be verified are sealed, the seal of the chapter to be verified can be determined to be complete.
In this embodiment, when determining whether the number of signature texts in the section to be verified is less than the number of signature identifiers in the target section and determining whether the number of seals in the section to be verified is less than the number of seal identifiers in the target section, it is also unnecessary to fill the signature texts and seal contents into the target dictionary to determine, and it is possible to directly determine whether the number of signature texts in the section to be verified is less than the number of signature identifiers in the target section and directly determine whether the number of seals in the section to be verified is less than the number of seal identifiers in the target section.
In this embodiment, after the signature text and the signature prompt text of the signature area and the seal content and the seal prompt text of the seal area are extracted, the computer device may further determine whether the signature text of the signature area is consistent with the signature prompt text of the signature area, and obtain a consistency determination result of the signature area, and further determine whether the seal content of the seal area is consistent with the seal prompt text of the seal area, and obtain a consistency determination result of the seal area.
For example, if the signature text of the signature area (i.e., the name signed by the principal) is "Zhang san" and the signature hint text of the signature area is "Liqu", then it may be determined that the signature text of the signature area is inconsistent with the signature hint text; the seal content of the seal area is Shenzhen xxx Limited company, and the seal prompt text of the seal area is Guangzhou xxx Limited company, so that the seal content of the seal area is inconsistent with the seal prompt text.
The consistency judgment of the signature area and the seal area can divide the signature area into two categories of 'signature text consistent with a signature prompt' and 'signature text inconsistent with a signature prompt', and the seal area into two categories of 'seal content consistent with a seal prompt' and 'seal content inconsistent with a seal prompt'.
The signature seal class image used for training the target detection model in step 202 includes various images such as an image with only a signature prompt, an image with only a seal prompt, an image with only a signature prompt and a signature, an image with only a seal prompt and a seal, an image with a signature prompt and a signature text and a seal prompt and a seal, an image with a signature prompt and a seal prompt, an image with a seal prompt and a seal prompt, and the like, and the classification result of the signature region includes various classification results such as only a signature prompt, only a signature prompt and a signature, a signature prompt and a name seal, only a signature, and the like, and the classification result of the seal region includes only a seal prompt, only a seal prompt and a seal, only a seal. Meanwhile, the consistency judgment result of the signature area is used as a classification result of the signature area, and the consistency judgment result of the seal area is used as a classification result of the seal area.
Then in step 202, the object detection model classifies the signature area and the seal area, and may divide each signature area and each seal area into a classification result. Therefore, the frequency of each classification result of all signature areas in the chapter to be verified and the frequency of each classification result of all seal areas can be counted, and the frequency counting result of the signature areas and the frequency counting result of the seal areas of the chapter to be verified are used as the signature of the chapter to be verified and the seal integrity verification result.
For example, the chapter to be verified has 5 signature areas which should be signed with a name and 5 seal areas which should be stamped, and 3 signature areas which are classified as "signed hint and signature and name seal" and 2 signature areas which are classified as "signed hint and signature only" are counted in the 5 signature areas; the 5 seal areas have 4 seal areas with the category of seal only prompt and seal, and 1 seal area with the category of seal only prompt. Therefore, the frequency counting result can be used as a signature seal integrity verification result of the chapter to be verified.
In this embodiment, the chapter to be verified may be any chapter in the document to be verified, or may be a chapter corresponding to a page image in which a signature and/or a seal exists in the document to be verified. Specifically, when judging whether a signature and/or a seal exists on a page, acquiring a target classification model which is trained in advance, wherein the target classification model is obtained by training a plurality of groups of page image training samples through a machine learning algorithm, each group of page image training samples comprises page images and label information for representing classification results of the page images, and inputting each page image of a document to be verified into the target classification model to obtain the classification results of each page image of the document to be verified, which are output by the target classification model, wherein the classification results comprise the page images with the signature and/or the seal and the page images without the signature and the seal. After the target classification model determines that the signature and/or the seal page image exist in the document to be verified, determining a chapter corresponding to the signature and/or the seal page image existing in the document to be verified as the chapter to be verified.
The machine learning algorithm used to train the object classification model may be a LeNet, alexNet, VGGNet, resNET, googleNet or other machine vision algorithm.
The embodiment also provides an implementation mode for verifying the seal. Specifically, after obtaining seal content of a seal area output by a text detection model, retrieving from a plurality of seal images stored in a database to obtain seal content and a standard seal image with seal type corresponding to seal content of a target seal area, calculating a similarity value of the seal image of the target seal area and the standard seal image according to an image similarity algorithm, and determining that the seal image of the target seal area does not pass through authenticity verification when the similarity value is smaller than a preset threshold value; if the similarity value is larger than a preset threshold value, determining that the seal image of the target seal area passes the authenticity verification.
When the target detection model determines the seal area in the chapter to be verified in step 202, the seal area may be further captured, so as to obtain a seal image of the seal area.
The seal types can be divided according to any standard, for example, the seal can be divided into a unit official seal and a personal name seal; the public chapters can be classified into national public enterprise chapters, private public enterprise chapters and other types according to the business property of the units. The division manner of the seal type is not limited.
The image similarity algorithm may be a similarity algorithm based on euclidean distance, or a similarity algorithm based on manhattan distance, or a cosine similarity algorithm. The cosine similarity algorithm represents the pictures as a vector, and the similarity of the two pictures is represented by calculating the cosine distance between the vectors corresponding to the two pictures respectively. The kind of the image similarity algorithm is not limited.
In addition, when the similarity value of the seal image of the target seal area and the standard seal image is smaller than the preset threshold value, the seal image of the target seal area is possibly changed, at this time, the computer equipment displays the seal image and the similarity value of the target seal area to a user, and the user judges whether the seal image of the target seal area is changed according to actual conditions. If the seal image is changed, the user can send an instruction to the computer equipment, instruct the computer equipment to store the seal image after the change of the target seal area, seal content and seal type corresponding to the target seal area in a database in an associated mode, and instruct the computer equipment to mark the update state of the seal image before the change of the target seal area as updated, so that the computer equipment can conveniently determine the updated seal image in subsequent operation. Wherein, can assign value to the updated state of the seal picture, different updated state assign different value. When the stamp image is never updated, the value of its update state may be null.
When the seal content and the standard seal image with the seal type corresponding to the seal content of the target seal area cannot be searched, the seal content and the seal type of the target seal area and the seal image of the target seal area are stored in a database in a correlated mode, so that the seal image of the target seal area in other documents can be verified later.
In the embodiment, the signature seal integrity of the document can be verified by filling the signature text and seal content into the target dictionary, so that the scheme can be realized.
The signature and seal integrity verification method in the embodiment of the present application is described above, and the following describes a computer device in the embodiment of the present application, referring to fig. 5, an embodiment of the computer device in the embodiment of the present application includes:
an obtaining unit 501, configured to obtain a current document to be verified, and obtain a pre-constructed target dictionary, where the target dictionary includes a correspondence between chapters of the document to be verified and signature identifiers and seal identifiers, and the target dictionary corresponds to a document category of the document to be verified;
The target detection unit 502 is configured to detect a document to be verified according to a target detection model that is trained in advance, and determine a signature area and a seal area in a chapter to be verified of the document to be verified;
A first determining unit 503, configured to determine a target chapter corresponding to a chapter to be verified in the target dictionary;
A judging unit 504, configured to judge whether the number of signature texts in all signature areas in the chapter to be verified is less than the number of signature identifiers in the target chapter, and judge whether the number of seals in all seal areas in the chapter to be verified is less than the number of seal identifiers in the target chapter;
A second determining unit 505, configured to determine that the signature and the seal of the chapter to be verified are complete when the number of signature texts is equal to the number of signature identifiers and the number of seals is equal to the number of seal identifiers;
the second determining unit 505 is further configured to determine that the signature of the chapter to be verified is incomplete when the number of signature texts is less than the number of signature identifiers;
The second determining unit 505 is further configured to determine that the stamp of the chapter to be verified is incomplete when the number of stamps is less than the number of stamp identifiers.
In a preferred implementation manner of this embodiment, the target detection unit 502 is specifically configured to obtain a target detection model, where the target detection model is obtained by training multiple sets of training samples by a machine learning algorithm, and each set of training samples includes a signature seal type image, where the signature seal type image includes an image with only a signature hint, an image with only a signature hint and a signature, an image with a signature text and a signature hint and a signature, an image with a signature hint and a signature, and an image with a seal hint and a signature hint; and inputting the document to be verified into the target detection model, and determining the region position information of the signature region and the region position information of the seal region in the chapter to be verified.
In a preferred implementation of this embodiment, the computer device further includes:
The text detection unit 506 is configured to obtain a text detection model, where the text detection model is obtained by training multiple sets of image training samples by using a text detection algorithm; inputting an image corresponding to the region position information of the signature region into a text detection model to obtain a signature text and/or a signature prompt text of the signature region output by the text detection model; and inputting an image corresponding to the region position information of the seal region into the text detection model to obtain seal content and/or seal prompt text of the seal region output by the text detection model.
In a preferred implementation manner of this embodiment, the signature identifier in the target chapter includes the region location information corresponding to the signature identifier, and the seal identifier in the target chapter includes the region location information corresponding to the seal identifier;
The computer device further comprises:
a filling unit 507, configured to fill the signature text of the signature area into a filling area indicated by a target signature identifier in the target chapter, where the area location information corresponding to the target signature identifier corresponds to the area location information of the signature area; and filling the seal content of the seal area into a filling area indicated by a target seal mark in the target chapter, wherein the area position information corresponding to the target seal mark corresponds to the area position information of the seal area.
In a preferred implementation manner of this embodiment, the determining unit 504 is specifically configured to determine whether a filling area indicated by all signature identifiers in the target chapter has a filling area of an unfilled signature text, and whether a filling area indicated by all seal identifiers in the target chapter has a filling area of an unfilled seal content;
The second determining unit 505 is specifically configured to determine that the signature of the chapter to be verified and the stamp are complete when there is no filled region of the unfilled signature text and there is no filled region of the unfilled stamp content; when a filling area of unfilled signature texts exists, determining that the number of signature texts is less than the number of signature identifiers, and the signature of the chapter to be verified is incomplete; when the filling area of unfilled seal content exists, the number of the seals is determined to be less than the number of the seal marks, and the seals of the chapter to be verified are incomplete.
In a preferred implementation manner of this embodiment, each set of training samples further includes tag information for representing a classification result of the signature seal class image;
the target detection unit 502 is specifically configured to input a document to be verified into the target detection model, so as to obtain a classification result of the signature area and a classification result of the seal area output by the target detection model.
In a preferred implementation manner of this embodiment, the determining unit 504 is further configured to determine whether the signature text of the signature area is consistent with the signature prompt text of the signature area, and obtain a consistency determination result of the signature area; judging whether the seal content of the seal area is consistent with the seal prompt text of the seal area, and acquiring a consistency judgment result of the seal area;
The computer device further comprises:
The statistics unit 508 is configured to count the frequency of each classification result of all signature areas in the section to be verified and the frequency of each classification result of all seal areas, where the classification results of the signature areas include a signature prompt only, a signature prompt and a signature only, a signature prompt and a name seal only, and a consistency judgment result of the signature areas only; the classification result of the seal area comprises consistency judgment results of only seal prompt, only seal prompt and seal, only seal and seal area; and taking the frequency statistics result of the signature area of the chapter to be verified as the signature of the chapter to be verified and the seal integrity verification result.
In a preferred implementation of this embodiment, the computer device further includes:
The target classification unit 509 is configured to obtain a target classification model that is trained in advance, where the target classification model is obtained by training multiple sets of page image training samples by a machine learning algorithm, and each set of page image training samples includes a page image and tag information that is used to represent a classification result of the page image; inputting each page image of the document to be verified into a target classification model to obtain a classification result of each page image of the document to be verified, which is output by the target classification model, wherein the classification result comprises a page image with a signature and/or a seal and a page image without the signature and the seal; and determining that a section to be verified corresponding to the page image of the signature and/or the seal exists in the document to be verified.
In a preferred implementation of this embodiment, the computer device further includes:
A seal verification unit 510, configured to retrieve seal content and a standard seal image corresponding to seal content of a target seal area, from a plurality of seal images stored in a database; calculating the similarity value of the seal image of the target seal area and the standard seal image according to an image similarity algorithm; when the similarity difference value is smaller than a preset threshold value, determining that the seal image of the target seal area does not pass the authenticity verification; and when the similarity value is larger than a preset threshold value, determining that the seal image of the target seal area passes the authenticity verification.
In a preferred implementation manner of this embodiment, the seal verification unit 510 is further configured to store, when the seal content and the standard seal image corresponding to the seal content of the target seal area cannot be retrieved, the seal content and the seal type of the target seal area and the seal image of the target seal area in association with each other.
In this embodiment, the operations performed by the units in the computer device are similar to those described in the embodiments shown in fig. 1 to 2, and are not repeated here.
In this embodiment, the target detection unit 502 detects the document to be verified according to the target detection model that is trained in advance, determines the signature areas and the seal areas in the section to be verified of the document to be verified, the judging unit 504 judges whether the number of signature texts in all signature areas in the section to be verified is less than the number of signature identifiers in the target section, and judges whether the number of seals in all seal areas in the section to be verified is less than the number of seal identifiers in the target section, and because the target dictionary is constructed by personnel according to the actual situation that the document needs to be signed and sealed, the target dictionary can be used as a standard for verifying the integrity of the signature seal of the document, and therefore, when the number of signature texts in all signature areas is less than the number of signature identifiers in the target section, the second determining unit 505 determines that the signature of the section to be verified is incomplete; when the number of the stamps in all the stamp areas is less than the number of the stamp marks in the target chapter, determining that the stamps of the chapter to be verified are incomplete. Therefore, the automatic verification of the integrity of the signature seal of the document is realized, the verification of the signature seal is not required to be manually executed, and the verification efficiency of the signature seal is improved.
Referring to fig. 6, an embodiment of a computer device according to the present application includes:
The computer device 600 may include one or more central processing units (central processing units, CPU) 601 and memory 605, where the memory 605 stores one or more application programs or data.
Wherein the memory 605 may be volatile storage or persistent storage. The program stored in the memory 605 may include one or more modules, each of which may include a series of instruction operations in the computer device. Still further, the central processor 601 may be arranged to communicate with the memory 605 to execute a series of instruction operations in the memory 605 on the computer device 600.
The computer device 600 may also include one or more power supplies 602, one or more wired or wireless network interfaces 603, one or more input/output interfaces 604, and/or one or more operating systems, such as Windows ServerTM, mac OS XTM, unixTM, linuxTM, freeBSDTM, etc.
The cpu 601 may perform the operations performed by the computer device in the embodiments shown in fig. 1 to 2, and detailed descriptions thereof are omitted herein.
The embodiment of the application also provides a computer storage medium, wherein one embodiment comprises: the computer storage medium has stored therein instructions which, when executed on a computer, cause the computer to perform the operations performed by the computer device in the embodiments of fig. 1-2 described above.
It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described systems, apparatuses and units may refer to corresponding procedures in the foregoing method embodiments, which are not repeated herein.
In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of the units is merely a logical function division, and there may be additional divisions when actually implemented, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other form.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in the embodiments of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application may be embodied essentially or in part or all of the technical solution or in part in the form of a software product stored in a storage medium, including instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a read-only memory (ROM), a random-access memory (RAM, random access memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
Claims (11)
1. A method of signature and seal integrity verification comprising:
acquiring a current document to be verified, and acquiring a pre-constructed target dictionary, wherein the target dictionary comprises the corresponding relation between chapters and sections of the document to be verified and signature identifiers and seal identifiers, and the target dictionary corresponds to the document category of the document to be verified;
Detecting the document to be verified according to a target detection model which is trained in advance, and determining a signature area and a seal area in a chapter to be verified of the document to be verified;
Determining a target chapter corresponding to the chapter to be verified in the target dictionary, judging whether the number of signature texts of all signature areas in the chapter to be verified is less than the number of signature identifiers in the target chapter, and judging whether the number of seals of all seal areas in the chapter to be verified is less than the number of seal identifiers in the target chapter;
If the number of the signature texts is equal to the number of the signature identifications and the number of the seals is equal to the number of the seal identifications, determining that the signature and the seal of the chapter to be verified are complete;
If the number of the signature texts is smaller than the number of the signature identifiers, determining that the signature of the chapter to be verified is incomplete;
And if the number of the seals is less than the number of the seal marks, determining that the seals of the chapter to be verified are incomplete.
2. The method for verifying signature and seal integrity as in claim 1 wherein the detecting the document to be verified according to a pre-trained target detection model comprises:
The target detection model is obtained by training a plurality of groups of training samples through a machine learning algorithm, each group of training samples comprises a signature seal type image, and the signature seal type image comprises an image with only a signature prompt, an image with only a seal prompt, an image with only a signature prompt and a signature, an image with only a seal prompt and a seal, an image with a signature prompt and a signature text and a seal prompt and a seal, an image with a signature prompt and a signature and a seal prompt, and an image with a seal prompt and a seal and a signature prompt;
and inputting the document to be verified into the target detection model, and determining the region position information of the signature region and the region position information of the seal region in the chapter to be verified.
3. The signature and seal integrity verification method of claim 2, further comprising:
acquiring a text detection model, wherein the text detection model is obtained by training a plurality of groups of image training samples through a text detection algorithm;
inputting an image corresponding to the region position information of the signature region into the text detection model to obtain a signature text and/or a signature prompt text of the signature region output by the text detection model;
And inputting an image corresponding to the region position information of the seal region into the text detection model to obtain seal content and/or seal prompt text of the seal region output by the text detection model.
4. The signature and seal integrity verification method as claimed in claim 3 wherein said signature identifier in said target chapter comprises region location information corresponding to said signature identifier, and said seal identifier in said target chapter comprises region location information corresponding to said seal identifier;
the method further comprises the steps of:
Filling the signature text of the signature region into a filling region indicated by a target signature identifier in the target chapter, wherein region position information corresponding to the target signature identifier corresponds to region position information of the signature region;
And filling the seal content of the seal area into a filling area indicated by a target seal mark in the target chapter, wherein the area position information corresponding to the target seal mark corresponds to the area position information of the seal area.
5. The method for verifying the signature and the seal integrity according to claim 4, wherein the determining whether the number of signature texts of all signature areas in the section to be verified is less than the number of signature identifications in the target section, and the determining whether the number of seals of all seal areas in the section to be verified is less than the number of seal identifications in the target section, comprises:
Judging whether filling areas indicated by all signature identifiers in the target section have filling areas of unfilled signature texts, and whether filling areas indicated by all seal identifiers in the target section have filling areas of unfilled seal contents;
If the filling area of the unfilled signature text does not exist and the filling area of the unfilled seal content does not exist, determining that the signature and the seal of the chapter to be verified are complete;
If a filling area of the unfilled signature text exists, determining that the number of the signature text is less than the number of the signature identifiers, and the signature of the section to be verified is incomplete;
If the filling area of the unfilled seal content exists, determining that the number of the seals is less than the number of the seal marks, and the seal of the section to be verified is incomplete.
6. A signature and seal integrity verification method as claimed in claim 3 wherein each set of said training samples further comprises tag information for representing a classification result of said signature seal class image;
The step of inputting the document to be verified into the target detection model, and determining the region position information of the signature region and the region position information of the seal region in the chapter to be verified comprises the following steps:
And inputting the document to be verified into the target detection model to obtain a classification result of the signature region and a classification result of the seal region, which are output by the target detection model.
7. The signature and seal integrity verification method as claimed in claim 6, further comprising:
Judging whether the signature text of the signature area is consistent with the signature prompt text of the signature area, and acquiring a consistency judgment result of the signature area;
Judging whether the seal content of the seal area is consistent with the seal prompt text of the seal area, and acquiring a consistency judgment result of the seal area;
Counting the frequency of each classification result of all signature areas and the frequency of each classification result of all seal areas in the section to be verified, wherein the classification results of the signature areas comprise consistency judgment results of only signature prompt, only signature prompt and signature, only signature prompt, signature and name seal, only signature and the signature areas; the classification result of the seal area comprises a seal prompt only, a seal only, and a consistency judgment result of the seal area only;
And taking the frequency statistics result of the signature area and the frequency statistics result of the seal area of the chapter to be verified as the signature and seal integrity verification result of the chapter to be verified.
8. A method of signature and seal integrity verification as claimed in any one of claims 1 to 7 wherein, prior to said detection of said document to be verified in accordance with a pre-trained target detection model, said method further comprises:
Obtaining a target classification model which is trained in advance, wherein the target classification model is obtained by training a plurality of groups of page image training samples through a machine learning algorithm, and each group of page image training samples comprises page images and label information used for representing classification results of the page images;
Inputting each page image of the document to be verified into the target classification model to obtain a classification result of each page image of the document to be verified, which is output by the target classification model, wherein the classification result comprises a page image with a signature and/or a seal and a page image without the signature and the seal;
and determining the section to be verified, corresponding to the page image of the signature and/or the seal, in the document to be verified.
9. The method of signature and seal integrity verification as claimed in claim 8 wherein after obtaining seal content and/or seal hint text for said seal area output by a text detection model, said method further comprises:
Retrieving seal content and standard seal images with seal types corresponding to seal content of a target seal area from a plurality of seal images stored in a database;
Calculating the similarity value of the seal image of the target seal area and the standard seal image according to the image similarity algorithm;
if the similarity value is smaller than a preset threshold value, determining that the seal image of the target seal area does not pass the authenticity verification;
and if the similarity value is larger than the preset threshold value, determining that the seal image of the target seal area passes the authenticity verification.
10. The signature and seal integrity verification method as claimed in claim 9, further comprising:
When the seal content and the standard seal image with the seal type corresponding to the seal content of the target seal area cannot be searched, the seal content and the seal type of the target seal area and the seal image of the target seal area are stored in the database in a correlated mode.
11. A computer device, comprising:
The acquisition unit is used for acquiring a current document to be verified and acquiring a pre-constructed target dictionary, wherein the target dictionary comprises the corresponding relation between chapters and signature identifiers and seal identifiers of the document to be verified, and the target dictionary corresponds to the document category of the document to be verified;
the target detection unit is used for detecting the document to be verified according to a target detection model which is trained in advance, and determining a signature area and a seal area in a chapter to be verified of the document to be verified;
a first determining unit, configured to determine a target chapter corresponding to the chapter to be verified in the target dictionary;
The judging unit is used for judging whether the number of signature texts of all signature areas in the chapter to be verified is less than the number of signature identifiers in the target chapter and judging whether the number of seals of all seal areas in the chapter to be verified is less than the number of seal identifiers in the target chapter;
The second determining unit is used for determining the signature and seal integrity of the chapter to be verified when the number of signature texts is equal to the number of signature identifiers and the number of seals is equal to the number of seal identifiers;
The second determining unit is further used for determining that the signature of the chapter to be verified is incomplete when the number of signature texts is smaller than the number of signature identifiers;
and the second determining unit is also used for determining that the seal of the chapter to be verified is incomplete when the number of the seals is less than the number of the seal marks.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011434668.9A CN112580108B (en) | 2020-12-10 | 2020-12-10 | Signature and seal integrity verification method and computer equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011434668.9A CN112580108B (en) | 2020-12-10 | 2020-12-10 | Signature and seal integrity verification method and computer equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112580108A CN112580108A (en) | 2021-03-30 |
| CN112580108B true CN112580108B (en) | 2024-04-19 |
Family
ID=75131140
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011434668.9A Active CN112580108B (en) | 2020-12-10 | 2020-12-10 | Signature and seal integrity verification method and computer equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112580108B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113434912B (en) * | 2021-08-24 | 2021-12-24 | 共道网络科技有限公司 | Material compliance verification method and device |
| CN114821624A (en) * | 2022-04-08 | 2022-07-29 | 烟台市勘察设计审查服务中心有限责任公司 | Signature authentication method and system for construction drawing examination |
| CN116092099B (en) * | 2023-01-18 | 2023-08-18 | 黑龙江省公安厅 | Multi-target administrative law enforcement document information integrity recognition detection method and system |
Citations (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB2104698A (en) * | 1981-04-27 | 1983-03-09 | Quest Automation | An improved apparatus for signature verification |
| US5606609A (en) * | 1994-09-19 | 1997-02-25 | Scientific-Atlanta | Electronic document verification system and method |
| US5992753A (en) * | 1997-10-30 | 1999-11-30 | Metanetics Corporation | Hand held dataform reader utilizing binarization process for dataform and signature area capture |
| CN101248433A (en) * | 2005-05-09 | 2008-08-20 | 普罗维拉公司 | Matching engine with signature generation and relevance detection |
| CN101325489A (en) * | 2007-06-15 | 2008-12-17 | 佳能株式会社 | Document verification method, document verification apparatus and storage medium |
| CN101360100A (en) * | 2008-09-16 | 2009-02-04 | 浙江汇信科技有限公司 | Digital signing, sealing and authenticating method for PDF document |
| CN106202568A (en) * | 2016-08-05 | 2016-12-07 | 欧阳能良 | Control system and standardized control method for medical laboratory quality management system file |
| CN107798301A (en) * | 2017-10-13 | 2018-03-13 | 上海眼控科技股份有限公司 | A kind of signature detection system and method for vehicle annual test |
| CN107818504A (en) * | 2017-09-27 | 2018-03-20 | 上海维信荟智金融科技有限公司 | The automation signature method and system of contract |
| CN109829269A (en) * | 2018-12-26 | 2019-05-31 | 平安科技（深圳）有限公司 | Method, apparatus and system based on E-seal authenticating electronic documents |
| CN110443596A (en) * | 2019-06-19 | 2019-11-12 | 深圳壹账通智能科技有限公司 | Data processing method, device, computer equipment and storage medium |
| CN110443773A (en) * | 2019-08-20 | 2019-11-12 | 江西博微新技术有限公司 | File and picture denoising method, server and storage medium based on seal identification |
| CN110443250A (en) * | 2019-07-31 | 2019-11-12 | 天津车之家数据信息技术有限公司 | A kind of classification recognition methods of contract seal, device and calculate equipment |
| CN110619274A (en) * | 2019-08-14 | 2019-12-27 | 深圳壹账通智能科技有限公司 | Identity verification method and device based on seal and signature and computer equipment |
| CN111144358A (en) * | 2019-12-30 | 2020-05-12 | 上海眼控科技股份有限公司 | Vehicle quality certificate verification method and device, computer equipment and storage medium |
| CN111160860A (en) * | 2019-12-27 | 2020-05-15 | 杭州日美开开电子科技有限公司 | File printing and stamping method |
| CN111368632A (en) * | 2019-12-27 | 2020-07-03 | 上海眼控科技股份有限公司 | Signature identification method and device |
| CN111401372A (en) * | 2019-12-20 | 2020-07-10 | 国家电网有限公司 | Method for extracting and identifying image-text information of scanned document |
| CN112001163A (en) * | 2020-09-03 | 2020-11-27 | 深圳证券信息有限公司 | Method and system for detecting integrity of file, electronic device and storage medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9729326B2 (en) * | 2008-04-25 | 2017-08-08 | Feng Lin | Document certification and authentication system |
-
2020
- 2020-12-10 CN CN202011434668.9A patent/CN112580108B/en active Active
Patent Citations (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB2104698A (en) * | 1981-04-27 | 1983-03-09 | Quest Automation | An improved apparatus for signature verification |
| US5606609A (en) * | 1994-09-19 | 1997-02-25 | Scientific-Atlanta | Electronic document verification system and method |
| US5992753A (en) * | 1997-10-30 | 1999-11-30 | Metanetics Corporation | Hand held dataform reader utilizing binarization process for dataform and signature area capture |
| CN101248433A (en) * | 2005-05-09 | 2008-08-20 | 普罗维拉公司 | Matching engine with signature generation and relevance detection |
| CN101325489A (en) * | 2007-06-15 | 2008-12-17 | 佳能株式会社 | Document verification method, document verification apparatus and storage medium |
| CN101360100A (en) * | 2008-09-16 | 2009-02-04 | 浙江汇信科技有限公司 | Digital signing, sealing and authenticating method for PDF document |
| CN106202568A (en) * | 2016-08-05 | 2016-12-07 | 欧阳能良 | Control system and standardized control method for medical laboratory quality management system file |
| CN107818504A (en) * | 2017-09-27 | 2018-03-20 | 上海维信荟智金融科技有限公司 | The automation signature method and system of contract |
| CN107798301A (en) * | 2017-10-13 | 2018-03-13 | 上海眼控科技股份有限公司 | A kind of signature detection system and method for vehicle annual test |
| CN109829269A (en) * | 2018-12-26 | 2019-05-31 | 平安科技（深圳）有限公司 | Method, apparatus and system based on E-seal authenticating electronic documents |
| CN110443596A (en) * | 2019-06-19 | 2019-11-12 | 深圳壹账通智能科技有限公司 | Data processing method, device, computer equipment and storage medium |
| CN110443250A (en) * | 2019-07-31 | 2019-11-12 | 天津车之家数据信息技术有限公司 | A kind of classification recognition methods of contract seal, device and calculate equipment |
| CN110619274A (en) * | 2019-08-14 | 2019-12-27 | 深圳壹账通智能科技有限公司 | Identity verification method and device based on seal and signature and computer equipment |
| CN110443773A (en) * | 2019-08-20 | 2019-11-12 | 江西博微新技术有限公司 | File and picture denoising method, server and storage medium based on seal identification |
| CN111401372A (en) * | 2019-12-20 | 2020-07-10 | 国家电网有限公司 | Method for extracting and identifying image-text information of scanned document |
| CN111160860A (en) * | 2019-12-27 | 2020-05-15 | 杭州日美开开电子科技有限公司 | File printing and stamping method |
| CN111368632A (en) * | 2019-12-27 | 2020-07-03 | 上海眼控科技股份有限公司 | Signature identification method and device |
| CN111144358A (en) * | 2019-12-30 | 2020-05-12 | 上海眼控科技股份有限公司 | Vehicle quality certificate verification method and device, computer equipment and storage medium |
| CN112001163A (en) * | 2020-09-03 | 2020-11-27 | 深圳证券信息有限公司 | Method and system for detecting integrity of file, electronic device and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 基于神经网络的发票文字检测与识别方法;蒋冲宇;鲁统伟;闵峰;熊寒颖;胡记伟;;武汉工程大学学报;20191215(第06期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112580108A (en) | 2021-03-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112580108B (en) | Signature and seal integrity verification method and computer equipment | |
| JP2014232533A (en) | System and method for ocr output verification | |
| WO2022126978A1 (en) | Invoice information extraction method and apparatus, computer device and storage medium | |
| CN111897962A (en) | Internet of things asset marking method and device | |
| CN113569863B (en) | Document checking method, system, electronic equipment and storage medium | |
| CN114359553B (en) | Signature positioning method and system based on Internet of things and storage medium | |
| CN114120299A (en) | Information acquisition method, device, storage medium and equipment | |
| CN111107107A (en) | Network behavior detection method and device, computer equipment and storage medium | |
| CN111552800A (en) | Abstract generation method and device, electronic equipment and medium | |
| CN115758451A (en) | Data labeling method, device, equipment and storage medium based on artificial intelligence | |
| CN104966109A (en) | Medical laboratory report image classification method and apparatus | |
| CN104899551B (en) | A kind of form image sorting technique | |
| CN112396047B (en) | Training sample generation method and device, computer equipment and storage medium | |
| CN111881900B (en) | Corpus generation method, corpus translation model training method, corpus translation model translation method, corpus translation device, corpus translation equipment and corpus translation medium | |
| CN115880702A (en) | Data processing method, device, equipment, program product and storage medium | |
| CN112989820B (en) | Legal document positioning method, device, equipment and storage medium | |
| CN116094971A (en) | Industrial control protocol identification method and device, electronic equipment and storage medium | |
| CN112949305B (en) | Negative feedback information acquisition method, device, equipment and storage medium | |
| CN113177543A (en) | Certificate identification method, device, equipment and storage medium | |
| CN114332599A (en) | Image recognition method, image recognition device, computer equipment, storage medium and product | |
| CN113888760A (en) | Violation information monitoring method, device, equipment and medium based on software application | |
| CN114049686A (en) | Signature recognition model training method and device and electronic equipment | |
| CN113724065B (en) | Auxiliary collecting method, device, equipment and storage medium based on flow guidance | |
| WO2023168964A1 (en) | Data segmentation method and related apparatus | |
| CN114925353A (en) | Account password resetting risk identification method, device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |